The evaluation and prediction of dam concrete quality have important theoretical significance for preventing catastrophes in concrete dams and predicting their long-term safety characteristics. To realize a comprehensive index for the evaluation and prediction of dam concrete characteristics, the fractal characteristics of dam concrete were analyzed using elastic-wave computed tomography (EWCT). The dam concrete quality evaluation model was established by combining the defined EWCT fractal dimensions with the strength and maturity of the dam concrete. Furthermore, a dam concrete artificial neural network (ANN) prediction model with a multi-index input layer was developed. The results indicated that the EWCT images exhibited the early evolution and distribution characteristics of dam concrete specimens, and there was a strong positive correlation between the EWCT fractal dimension and the age of the dam concrete. The grey relational analysis coefficient between the EWCT fractal dimension and fully graded concrete strength was similar to the coefficient between the wet-screened and fully graded concretes. Among the five dam concrete detection indices used in the ANN, the weight contribution rate of the EWCT fractal dimension to the dam concrete ANN prediction model was 23.15 %. In conclusion, the fractal dimension may be used as a critical index in dam concrete detection to pave the way for developing intelligent dam concrete detection and monitoring based on non-destructive technology.